Multi-objective optimization of best management practices at watershed scale: A case study of drinking water source watersheds in northeast black soil region of China

The Northeast Black Soil Region, a key grain-producing area in China, faces severe gully erosion and water eutrophication due to decades of intensive farming and fertilizer use. Numerous best management practices (BMPs) have been implemented to control agricultural diffuse pollution in this area, but few studies have quantitatively evaluated the effectiveness and spatial optimization of these BMPs at the watershed scale. This study focuses on the upstream watershed of Shitoukoumen reservoir, the drinking water source for Changchun. The Soil and Water Assessment Tool (SWAT) model was first used to assess the pollution load reduction efficiency of four typical BMPs-Residue Cover tillage (RC), Grassed Waterway (GW), Vegetative Filter Strip (VFS), and Check Dam (CD)-at the sub-watershed scale. The Non-dominated Sorting Genetic Algorithm II (NSGA-II) was then applied to explore optimal BMPs spatial configurations under various constraints. The evaluation results showed that among individual BMPs, VFS had the highest pollutant removal efficiency, while CD had the lowest nutrient load reduction capacity. Combining BMPs, especially source and end-point controls, significantly improved sediment and phosphorus reduction. When considering multiple constraints such as pollution load reduction, implementation costs, and land occupation, the optimized BMPs spatial configuration could reduce sediment by over 50 %, total nitrogen (TN) by over 30 %, and total phosphorus (TP) by over 40 %, while increasing farmland by over 800 ha, with a total investment of around 1.1 billion RMB. Moreover, although the BMPs spatial optimization scheme considering all three constraints is the optimal solution for diffuse pollution control in the study area, its implementation is relatively complex. The investment required is more than 10 times higher than that of the least expensive option, yet the improvement in pollution reduction efficiency does not increase proportionally. Therefore, future BMPs optimization goals in the black soil region should consider pollution severity, available funding, and farmer acceptance to select the most suitable local solutions.